This proposal seeks to clarify two central issues in the field of grapheme-color synesthesia. Grapheme-color synesthetes experience consistent and automatic color percepts when viewing or thinking of graphemes such as letters or numbers. The central issues are 1) whether grapheme-color synesthesia relies on direct anatomical projections or indirect functional connectivity between cortical regions and 2) to what extent synesthetic percepts require attention. A debate exists amongst synesthesia researchers, some of whom suggest that synesthesia results from incomplete cortical pruning during early development (e.g., Maurer & Mondloch, 2005; Ramachandran & Hubbard, 2001) and others who suggest it results from disinhibition of connections normally suppressed in all adults (e.g., Grossenbacher & Lovelace, 2001). A growing body of convergent functional and anatomical evidence supports the involvement of color-selective visual area V4 (Hubbard, Arman, Ramachandran, & Boynton, 2005); Weiss & Fink, 2008) and also inferior parietal areas (intraparietal sulcus and angular gyrus, e.g., (Rouw & Scholte, 2007; Nunn et al., 2002). Likewise, TMS to the angular gyrus reliably diminishes synesthetic percepts (Esterman, Verstynen, Ivry, & Robertson, 2006; Muggleton, Tsakanikos, Walsh, & Ward, 2007). This growing corpus of evidence has consistently identified both parietal cortex and area V4 in synesthetic perception but no study has clarified how these two areas might interact if at all. To understand synesthesia and to resolve the ongoing debate of extra connections or disinherited feedback (or both) it is necessary to map the functional circuit involved during synesthetic experience and co-register it with accurate measurements of retinotopically defined areas of visual cortex and anatomical connectivity (DTI). We will retinotopically localize visual areas from V1 through IPS and measure both anatomical and functional connectivity within these regions. Then, using a psychophysical paradigm developed from previous work on synesthesia, we will test the cortical networks involved in grapheme-color synesthesia and quantitatively measure their dependence on attention. More generally, this proposal addresses the broad issues of perceptual binding and feature integration, core issues in the field of perception and cognitive neuroscience and psychology. Synesthetic percepts do not replace their inducers and instead are added to them as an extra feature. This synesthetic concurrent is also most likely mediated by attention (see Mattingley, 2009 for a review). As suggested by Robertson (2003), synesthesia offers an ideal model for experimentally testing and isolating perceptual binding mechanisms and feature integration because a subjective synesthetic experience can be measured as a purely isolated phenomenological experience of a single feature (i.e., synesthetic color) while all stimulus properties (i.e., the stimulus luminance, color, size, object type, etc.) remain measurably constant.